Sindre Benjamin Remman
Papers
3
Total Citations
29
H-Index
2
About
Sindre Benjamin Remman is a researcher at the intersection of robotics and explainable artificial intelligence (XAI), focusing on making autonomous systems more transparent and trustworthy. His primary contributions lie in developing and analyzing interpretable deep reinforcement learning (DRL) methods for robotic control. In his highly cited 2021 work (21 citations), he demonstrated how to train a robotic manipulator to operate a lever using DRL with Hindsight Experience Replay, while simultaneously generating Shapley value-based explanations for the policy's decisions. Remman further advanced this field by investigating the critical distinction between causal and marginal Shapley values (2022, 7 citations), showing how incorporating causal knowledge about a robot's state relationships can produce more meaningful explanations. His most recent work (2025) introduces a novel method for generating realistic counterfactual explanations for mobile robots controlled by machine learning, using 2D LiDAR data to help researchers understand what would need to change for a robot to make a different decision. Remman’s research is essential for building safer, more interpretable autonomous systems that can be deployed with confidence.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3